Forecasting foreign currency exchange rates for department of defense budgeting1
| Published date | 01 April 2017 |
| Pages | 315-336 |
| Date | 01 April 2017 |
| DOI | https://doi.org/10.1108/JOPP-17-03-2017-B002 |
| Author | Nicholas R. Gardner,Jonathan D. Ritschel,Edward D. White,Andrew T. Wallen |
| Subject Matter | Public policy & environmental management,Politics,Public adminstration & management,Government,Economics,Public Finance/economics,Texation/public revenue |
JOURNAL OF PUBLIC PROCUREMENT, VOLUME 17, ISSUE 3, 315-336 FALL 2017
FORECASTING FOREIGN CURRENCY EXCHANGE RATES FOR
DEPARTMENT OF DEFENSE BUDGETING1
Nicholas R. Gardner, Jonathan D. Ritschel,
Edward D. White, and Andrew T. Wallen*
ABSTRACT. This paper examines the opportunity cost of applying s imple
averages in formulating the Department of Defense (DoD) budget for foreign
exchange rates. Using out-of-sample validation, we evaluate the status quo
of a center-weighted average against a Random Walk model, ARIMA, forward
rates, futures contracts, and a private firm’s forecasts over two time periods
extending from Fiscal Year (FY) 1991 to FY 2014. The results strongly
indicate that four of the alternative methods outperform the status quo over
the shorter time period, and three methods for both time periods.
Furthermore, a non-parametric comparison of the median error demonstrates
statistical similarities between the four alternative methods over the short
term. Overall, the paper recommends using the futures option prices to
decrease forecast error by 3.23% and avoiding a $34 million opportunity cost.
INTRODUCTION
Uncertainty engulfs the resource planning process to meet
expected needs as individuals, companies, and governments each
face this dilemma in deciding how best to allot resources to their
requirements. Many vested-interests complicate a government’s
budget allocation while the process’s complexity increases when
government budgets for those requirements in the home currency and
expense for those requirements in a foreign currency. The ambiguity
--------------------------------------
* Nicholas R. Gardner, M.S., is a cost analyst with the US Air Force. Jonathan
D. Ritschel, Ph. D., is Director, Cost Analysis Graduate Program at the Ai r
Force Institute of Technology. Andrew T. Wallen, Ph. D., is the Chief
Economist for the Secretary of the Air Force, Directorate of Economics and
Business Management. Edward D. White, Ph. D., is a Professor of Statistics at
the Air Force Institute of Technology. Their research interests include public
procurement and DoD acquisitions.
Copyright © 2017 by PrAcademics Press
316 GARDNER, RITSCHEL, WHITE & WALLEN
of the future exchange rate between the two can induce further
variability resulting in a resource allocation inconsistent with resource
needs. Inconsistency between an allocation and needs results in an
opportunity cost as other requirements lose budget authority from
inaccurate exchange rate estimates. This paper examines current DoD
foreign exchange rate forecasting practices and analyzes five
alternatives methods: a Random Walk, ARIMA, Forward Rates,
Futures, and a Private Firm’s Forecasts.
The DoD operates in every time zone and in every climate around
the globe at more than 5,000 different locations or sites. In executing
the US strategy across the globe, the DoD incurs requirements
denoted in local currencies. For FY 2013, the DoD budgeted $5.5
billion for foreign currency with an increase to $5.6 billion in FY2014.
In order to reduce the variability between the estimated and actual
exchange rate, Congress authorized a foreign currency fluctuation
(FCF) account in FY79 (Department of Defense, 2011). The DoD uses
this account to disburse additional funds when the exchange rate
budgeted is insufficient or collect excess funds for budgeted
surpluses. The FCF held $1.1B for FY2013 and $1.0B for FY2014.
The FCF account represents the opportunity cost of inaccurate
estimates.
In order to reduce the size of the FCF account, this paper reviews
the current DoD forecasting process as well as the academic literature
on exchange rate forecasting. A vast literature on foreign exchange
rate forecasting exists, but we will only touch on some promising
methods due to the limitations in the federal budget process (e.g. the
timing between forecasting a rate and the passing of a federal budget
or computational requirements). Exploring options available in the
private sector provides additional context to the DoD’s procedures.
Next, we review five methods from the literature to compare against
the status quo. These methods were chosen based on applicability to
the federal budget process, ease of use, and academic research.
Each method compares a forecasted budgeted rate to the actual rate
by the absolute percent error (APE) between the two rates. An APE
allows for comparison between the different currencies’ values (e.g.
100 Yen/US Dollar versus 0.6 Euro/US Dollar). The methods are then
evaluated against each other and the status quo before scrutinizing
whether each method is statistically different than the other. The
To continue reading
Request your trial